[00:00:06] Speaker 04:
Our next case this morning is number 18-1863, Coney Delight, PNNV v. Jamalto, M2M.

[00:01:30] Speaker 04:
Okay, Mr. Healy.

[00:01:39] Speaker 02:
How do you pronounce the plaintiff's name?

[00:01:41] Speaker 02:
Is it Koninkliky?

[00:01:43] Speaker 00:
That's a tough one, Your Honor.

[00:01:45] Speaker 00:
Koninkliky.

[00:01:47] Speaker 00:
But we just go by KPN.

[00:01:50] Speaker 02:
What did you say before, KPN?

[00:01:52] Speaker 00:
Koninkliky.

[00:01:54] Speaker 02:
Oh, the J counts.

[00:01:56] Speaker 00:
The client representative is actually here and speaks Dutch, so he may be able to help us if you have

[00:02:00] Speaker 00:
But that's how we do it.

[00:02:02] Speaker 00:
KPN is probably the easiest way for the court.

[00:02:05] Speaker 00:
We'll go with that.

[00:02:07] Speaker 00:
Thank you, Your Honors.

[00:02:09] Speaker 00:
So good morning, Your Honors.

[00:02:10] Speaker 00:
May it please the court.

[00:02:10] Speaker 00:
My name's Andres Healy.

[00:02:12] Speaker 00:
I'm here today on behalf of KPN.

[00:02:14] Speaker 00:
And for at least three reasons, the district court below.

[00:02:17] Speaker 04:
I don't think you guys did a great job in explaining this technology either side, which is a problem in these cases.

[00:02:24] Speaker 04:
So what is the advantage?

[00:02:30] Speaker 04:
of using scrambled original data here is to do the checking.

[00:02:38] Speaker 00:
What's the advantage of that?

[00:02:40] Speaker 00:
Absolutely, Your Honor.

[00:02:41] Speaker 00:
Apologies for that.

[00:02:42] Speaker 00:
The advantage here is set forth specifically in the specification.

[00:02:46] Speaker 00:
For example, if you look in the summary of the invention, this is appendix 71, column 2, lines 48 through 58.

[00:02:53] Speaker 00:
It expressly says, the invention is therefore based on the insight

[00:02:58] Speaker 00:
that an error which repeats itself, once not detected by a normal fixed checking function, will repeatedly go undetected.

[00:03:06] Speaker 01:
How do you, can you explain, though, in basic terms, why it is that permutating the data would solve that problem?

[00:03:15] Speaker 00:
Sure.

[00:03:16] Speaker 00:
So in the prior art, Your Honors, the data is operated upon by the generating device.

[00:03:22] Speaker 00:
It's encoded, for lack of a better word.

[00:03:24] Speaker 00:
It's then transmitted to effectively the receiver.

[00:03:28] Speaker 00:
If there's some sort of electromagnetic interference... You're repeating the same check under the prior art, right?

[00:03:34] Speaker 00:
Correct, Your Honor.

[00:03:35] Speaker 00:
And the problem with this is that there is some sort of systematic electromagnetic interference, for example.

[00:03:41] Speaker 00:
So there's a microwave in the office building.

[00:03:43] Speaker 00:
There's a TV.

[00:03:45] Speaker 00:
I guess the TV is a better example.

[00:03:47] Speaker 00:
It's always on.

[00:03:48] Speaker 00:
So that electromagnetic radiation hits that transmitted sequence in the same spot every time.

[00:03:54] Speaker 00:
So for example, if you have a sequence of 10 bits, it always hits bits two and eight.

[00:03:59] Speaker 00:
And it always hits those two bits.

[00:04:02] Speaker 00:
If you're always using the same encoding process, it's always going to hit the same two bits.

[00:04:06] Speaker 00:
The receiving end is never going to know there's a problem.

[00:04:09] Speaker 00:
If you are changing the bit position of those 10 bits.

[00:04:13] Speaker 01:
Why would the receiving end not know that there's a problem if there's just two bits, two and eight?

[00:04:19] Speaker 01:
Can you explain how that would happen?

[00:04:22] Speaker 00:
why I would not know that there was a problem.

[00:04:23] Speaker 00:
Because it's going to happen every single time.

[00:04:25] Speaker 00:
So for example, if you have a device, for example, your refrigerator that's always, or your fire alarm is a good example.

[00:04:31] Speaker 00:
It's always pinging the network, no fire, no fire, no fire, no fire.

[00:04:35] Speaker 00:
If it always has the same error at each point, it's always going to get the same message.

[00:04:40] Speaker 00:
It's never going to detect that there's a problem.

[00:04:42] Speaker 00:
If you have, in contrast, in the innovation plan in the 662 patent, you now introduce this varying component to the computer.

[00:04:52] Speaker 00:
The varying component says you take the data, you change the bit position of some of the bits, and then in addition to just... You use different check data for different transmissions, right?

[00:05:03] Speaker 00:
You can.

[00:05:03] Speaker 00:
That is certainly the component set forth in the dependent claims where the... Which are all involved here.

[00:05:08] Speaker 04:
It's two, three, and four.

[00:05:09] Speaker 00:
Correct, Your Honor.

[00:05:10] Speaker 00:
We disclaim claim one.

[00:05:11] Speaker 00:
I would say certainly for the point of the purpose of the record, claim one, in my opinion, and I think the claim's buried out,

[00:05:18] Speaker 00:
sets forth the overall architecture.

[00:05:20] Speaker 00:
It sets forth and describes the input, which is the binary data.

[00:05:24] Speaker 00:
It sets forth the order of operations of the various computer components.

[00:05:28] Speaker 00:
The varying device has to act first by permutating the bit position of at least some of the bits.

[00:05:33] Speaker 00:
It then has to provide that now varied data to the generating device.

[00:05:37] Speaker 04:
OK, but she didn't really answer my earlier question, which is why scrambling the original data, which is what these claims require, is beneficial, as opposed to

[00:05:47] Speaker 04:
generating new data, which the specification recognizes as an alternative.

[00:05:53] Speaker 00:
Correct.

[00:05:53] Speaker 00:
So again, my apologies, Your Honor.

[00:05:55] Speaker 00:
With respect to that specific question, the process that's contemplated in the prior art, just a bit of background, is that you send not only the encoded data, so the supplementary check data, but you send the original data as well.

[00:06:08] Speaker 00:
And then that goes and it's received by the receiver, which does a comparison.

[00:06:12] Speaker 00:
So if you always have this same formula going through,

[00:06:15] Speaker 00:
and the electromagnetic interference hits them at the same point, that comparison check can result.

[00:06:21] Speaker 00:
I mean, in many cases, it would result.

[00:06:23] Speaker 04:
That happens if you use the same data all the time.

[00:06:26] Speaker 04:
But why is it beneficial to scramble the original data as opposed to using new data?

[00:06:31] Speaker 00:
Right.

[00:06:31] Speaker 00:
And so the point here, Your Honor, is the scrambling of the data

[00:06:39] Speaker 00:
you take the positions, the bit positions of at least some of those bits, you reorder them.

[00:06:43] Speaker 00:
So now when you're sending the data across the stream, you have the original data.

[00:06:48] Speaker 00:
You have now the reorder data.

[00:06:52] Speaker 00:
When that, again, assuming that there's this electromagnetic interference, it hits it at the same point.

[00:06:57] Speaker 00:
So in both sets of data, it interferes.

[00:06:59] Speaker 04:
I'm sorry.

[00:06:59] Speaker 04:
I'm not understanding why you're saying why, what your explanation is, is to why scrambling original data is more beneficial than using new data.

[00:07:11] Speaker 01:
Is it because on the other end of the system, when you run it through the original data that you sent has to be run through the function again to create the modified data that's compared to the modified data that was sent from the beginning?

[00:07:27] Speaker 01:
Is that why you wouldn't want to just send new data at the beginning of the system?

[00:07:32] Speaker 01:
Do you understand the question that's being asked?

[00:07:35] Speaker 00:
Apologies.

[00:07:35] Speaker 00:
I'm not sure I understand the sending of the new data.

[00:07:39] Speaker 04:
The specification talks about a random data generator as an alternative to scrambling the original data, correct?

[00:07:47] Speaker 00:
Yes, your honor.

[00:07:48] Speaker 04:
What is the benefit of scrambling the original data as opposed to using the alternative of generating new data?

[00:07:58] Speaker 00:
When you say generating new data, you're referring to the random number generator?

[00:08:01] Speaker 00:
Yeah.

[00:08:02] Speaker 00:
The benefit is that the random number generator, for example, if you're using the parity bit function example, would be, if you have an odd number of bits, the number that's added to the end would be a 1.

[00:08:13] Speaker 00:
If you have an even number of bits, it would be a 2.

[00:08:16] Speaker 00:
It's a more limited checking function.

[00:08:19] Speaker 00:
It allows for less capable checking.

[00:08:24] Speaker 00:
Varying of the bits.

[00:08:25] Speaker 00:
So for example, you have 10-bit stream.

[00:08:27] Speaker 00:
You change the bit position of bits 2 and 8.

[00:08:30] Speaker 00:
You send that generated.

[00:08:32] Speaker 00:
That's the supplementary check data.

[00:08:33] Speaker 00:
You send that across.

[00:08:34] Speaker 00:
It goes across with the supplementary or with the original data.

[00:08:38] Speaker 00:
At the receiving end, it can now compare it.

[00:08:40] Speaker 00:
It's going to go through the reverse of that process as contemplated in the process that existed in the prior art.

[00:08:45] Speaker 00:
And so now, if you, for example, you had that electromagnetic radiation and hit both streams, the original data,

[00:08:52] Speaker 00:
and the check data at position two and eight, because you've reordered those bits when they become, you know, they're put back into their original formation, they're different.

[00:09:02] Speaker 00:
So there is now a much better way of checking because you have, you know, the scrambling of those bit positions effectively moves the bits and means that the electromagnetic radiation no longer hits them in the same spot.

[00:09:16] Speaker 04:
It may be me, but I'm not understanding what you're saying.

[00:09:20] Speaker 04:
If I understood correctly what you're saying, you're comparing the benefits of using scrambling data as opposed to the original data, which was not my question.

[00:09:31] Speaker 04:
Let me put it differently.

[00:09:32] Speaker 04:
Do these claims cover every way of checking for systemic errors that don't use the original data?

[00:09:44] Speaker 00:
No.

[00:09:46] Speaker 04:
And why not?

[00:09:47] Speaker 00:
Well, for example, the prior art contemplates the parity bit method, which is not something that would be contained within the claims as set forth in the 662 patent.

[00:09:57] Speaker 00:
There's also the prior art contemplates the CRC method, also outside the claims.

[00:10:03] Speaker 00:
Frankly, the example your honor gave of Figure 2, the random number generator, not included within the scope of the 662 patent claims.

[00:10:11] Speaker 00:
An even more simplistic example is

[00:10:13] Speaker 00:
The claims of the 662 patent, particularly claim one, again, sets forth a very particular input, output, and an order of operations to get from that input to output.

[00:10:24] Speaker 01:
Do I understand correctly you're not relying on claim one in an appeal anymore?

[00:10:27] Speaker 01:
You are just relying on claims two, three, and four?

[00:10:31] Speaker 01:
I mean, you're not focusing on those.

[00:10:33] Speaker 01:
Do I understand that?

[00:10:35] Speaker 00:
We are not relying on claim one as being independently, well, I should say, we're not relying on claim one because of the fact that we have disclaimed it.

[00:10:43] Speaker 00:
I still think claim one is important because it sets forth the structure and the architecture of this claim and provides the context.

[00:10:49] Speaker 04:
But we don't have to decide if claim one is one-on-one eligible.

[00:10:52] Speaker 00:
We do not, Your Honor.

[00:10:53] Speaker 00:
However, again, obviously the dependent claims all depend from claim one, and so it provides context.

[00:10:58] Speaker 00:
Just to finish answering your question, Your Honor, with respect to, you know, does this preempt effectively all implementations of this?

[00:11:05] Speaker 00:
again the simplest example is if you were to just Move the structure the organization of the of the the components you had the generating device operate first And then the interleaving device the permutating device operate also not covered because again the claim one architecture requires specific order of operations and

[00:11:26] Speaker 00:
So hopefully that answers your question, Your Honor.

[00:11:28] Speaker 02:
Can I see if I can try to articulate an understanding of how this claimed invention works and maybe why there's an advantage?

[00:11:36] Speaker 02:
I mean, I think I agree with Judge Dyke that your brief proclaims over and over and over again, this is an advantage.

[00:11:45] Speaker 02:
This detects systemic errors.

[00:11:48] Speaker 02:
But it doesn't quite piece together why that is true by using permutation and then varying the permutation over time.

[00:11:56] Speaker 02:
It has an abstract quality.

[00:11:59] Speaker 02:
The brief.

[00:12:00] Speaker 02:
The brief.

[00:12:01] Speaker 02:
Maybe the claims too, but at least the brief.

[00:12:03] Speaker 00:
Thank you.

[00:12:04] Speaker 02:
So let's try to bring it down to brass tacks.

[00:12:08] Speaker 02:
Sure.

[00:12:08] Speaker 02:
The problem is that is with checked data.

[00:12:12] Speaker 02:
Checked data is supposed to be some representative value of the content of the data inside a given data block.

[00:12:21] Speaker 02:
But because it's just a shorthand value, shorthand label,

[00:12:25] Speaker 02:
for all of that data in the data block, it can't precisely describe all of the data, because it's very shorthand.

[00:12:35] Speaker 02:
And so what that means is you could have a given checked data value that corresponds to not only that particular data inside that given data block, but it also might correspond to other versions of data.

[00:12:52] Speaker 02:
And so therefore, you can come across a problem in the data transmission where a given data block is corrupted in a particular way, at particular bits, but still nevertheless generate on the back end, on the receiving end, the exact same check data value.

[00:13:11] Speaker 02:
And so to the error detection scheme, it will look like the data block has been transmitted cleanly because the checked data values on the front end and the back end are equal to each other.

[00:13:24] Speaker 02:
And so if it's a systemic error, then what's going to happen is throughout your entire data transmission, block after block after block, you're going to have the exact same error, and the detection scheme is not going to be able to figure that out.

[00:13:38] Speaker 02:
because of the limitations of checked data values, right?

[00:13:43] Speaker 00:
Correct.

[00:13:44] Speaker 02:
OK.

[00:13:44] Speaker 02:
So what you've done is you've decided you're going to permutate the data, the original data in each given data block.

[00:13:53] Speaker 02:
And then you're going to vary that permutation, maybe not for every single block, but for several blocks in a given stream of data block transmissions.

[00:14:03] Speaker 02:
And by varying the permutation,

[00:14:06] Speaker 02:
even if the first permutation doesn't capture and reveal the persistent error, because it still ends up generating the exact same check.

[00:14:20] Speaker 02:
data and value on the back end.

[00:14:22] Speaker 02:
By varying the permutation at other given data blocks, you're eventually going to come up with a variance in the data that will reveal the error and not keep it concealed.

[00:14:38] Speaker 02:
So at some point in time, you're going to figure out, aha, we've got an error here.

[00:14:44] Speaker 02:
Is that how the claimed invention works?

[00:14:46] Speaker 00:
I would say at a high level, yes, Your Honor.

[00:14:48] Speaker 00:
And I would say that the examples you gave certainly get into some of the further down the stream of the dependent claims, but in a general sense, yes.

[00:14:54] Speaker 00:
Right.

[00:14:55] Speaker 02:
Well, when you say at a high level, I mean, to me, that's a lower level than whatever level your brief was at.

[00:15:01] Speaker 00:
Understood.

[00:15:01] Speaker 00:
Understood.

[00:15:02] Speaker 00:
OK.

[00:15:03] Speaker 02:
So can you tell me what's going on with claim three?

[00:15:09] Speaker 02:
Claim two, you have a theory.

[00:15:11] Speaker 02:
where you're varying the permutation over time.

[00:15:15] Speaker 02:
And I assume that means for successive data blocks, you're using different permutations, different methodologies of switching the bits around, scrambling the data.

[00:15:28] Speaker 02:
But I don't understand claim three.

[00:15:30] Speaker 00:
So claim two, Your Honor, just the claim construction is from time to time.

[00:15:34] Speaker 00:
And yes, it would allow for the capability to change the permutation that the varying device is applying.

[00:15:40] Speaker 00:
Claim three is that you're changing that permutation based on the original data So as described in the specification that can give me an example sure the number of bits is the simplest example if the Input data block has 50 bits it receives this permutation if it has 100 bits it receives a different permutation

[00:16:00] Speaker 00:
The specification, I think this is at columns 5 and 6, also describes other examples.

[00:16:07] Speaker 00:
For example, it can be based on an index.

[00:16:09] Speaker 00:
It can be based on other qualities of the data itself.

[00:16:13] Speaker 02:
Why would that matter?

[00:16:15] Speaker 02:
Oh, if it's 20 bits, use this permutation.

[00:16:17] Speaker 02:
If it's 50 bits, use that permutation.

[00:16:21] Speaker 02:
Why would that be an inventive concept?

[00:16:25] Speaker 00:
Why would it be an inventive concept?

[00:16:28] Speaker 00:
for us to why it would be an invented concept, I would rely on the spec.

[00:16:32] Speaker 00:
And just to take one step back, a lot of the questions about the utility of this invention, does it actually do what's promised?

[00:16:38] Speaker 00:
Does it actually confer the benefit that it states?

[00:16:42] Speaker 02:
So what benefit is conferred by using a different permutation based on the length of bits in a given data block?

[00:16:48] Speaker 02:
What's the benefit?

[00:16:50] Speaker 02:
What's the technical benefit?

[00:16:53] Speaker 00:
I mean, number one, the technical benefit would be, and again, the entire context of what we're claiming is a component of it.

[00:16:58] Speaker 04:
But isn't the simple answer that if

[00:17:02] Speaker 04:
claim 2 is patent eligible, since claim 3 incorporates claim 2 as a dependent claim, it's patent eligible too, even if it doesn't add anything that's an inventive concept to claim 2.

[00:17:15] Speaker 00:
Right?

[00:17:16] Speaker 00:
Absolutely agree with that, Your Honor.

[00:17:17] Speaker 02:
But your brief attempted to argue there was an independent reason for why claim 3, aside from claim 2,

[00:17:26] Speaker 02:
Is patent eligible and I don't understand it right now and you've given me you've got 15 seconds to tell me What's the answer sure again the overall system contemplates that there be?

[00:17:38] Speaker 01:
Both a receiving end and a transmitting end and you know again What's the inventive concept that claim three adds to claim two assuming?

[00:17:45] Speaker 01:
Let's assume for a minute that I agreed with you on claim two and

[00:17:50] Speaker 01:
then you have nothing else to add for claim three.

[00:17:53] Speaker 01:
But if I don't agree with you on claim two, how are you going to convince me that claim three adds something that makes it an inventive concept?

[00:18:01] Speaker 00:
Sure.

[00:18:01] Speaker 00:
And I apologize.

[00:18:02] Speaker 00:
I'm trying to answer that.

[00:18:03] Speaker 01:
It's the same question.

[00:18:04] Speaker 00:
I'm asking you the same thing.

[00:18:05] Speaker 00:
Of course.

[00:18:06] Speaker 00:
And so the idea that you're basing it on the characteristics of the original data, where that comes in and where that confers the benefit is because you have both the receiving and the transmitting end.

[00:18:16] Speaker 00:
When you're basing your permutations on the characteristics of the original data,

[00:18:20] Speaker 00:
That allows for there to be a similarity and for both the receiving and transmitting end because it's based on the characteristics of the data.

[00:18:27] Speaker 00:
The data is going to be the same at both ends.

[00:18:30] Speaker 00:
That is where the benefit comes in.

[00:18:31] Speaker 00:
You now are allowing for what the patent recites and the specification describes as that added benefit conferred when both ends are using the same permutation structure and the same permutation point.

[00:18:42] Speaker 01:
But that's satisfied by claim one or claim two, that they're using the same permutation device at the beginning and the end.

[00:18:49] Speaker 01:
Claim three is just talking about how the permutation device itself is going to be based on the original data.

[00:18:55] Speaker 00:
The permutation applied is based on the original data in claim three.

[00:18:59] Speaker 02:
So wait a second.

[00:19:00] Speaker 02:
You're telling me claim three is specifically calling for the same permutation to be done at the sending end and the receiving end.

[00:19:08] Speaker 02:
That's all claim three is calling for?

[00:19:10] Speaker 00:
I'm saying that claim three requires that the permutation applied be based on the characteristics of the original data.

[00:19:16] Speaker 00:
One of the examples of that is if you, for example, have a block that has a set number of bits, because now you're determining what permutation formula, in the manner in which you're going to permutate based on those characteristics, now you're incorporating and bringing in the fact that both the receiving and the transmitting end are going to apply the same formula or can apply the same permutation formula to those bits.

[00:19:39] Speaker 02:
Would you want to always use the same permutation formula when you're doing this kind of error check?

[00:19:44] Speaker 00:
I think that would be the best case scenario, yes, your honor, which is why we think Claim 3 introduces and maintains that inventive concept.

[00:19:54] Speaker 04:
All right.

[00:19:55] Speaker 04:
Anything more?

[00:19:57] Speaker 04:
All right.

[00:19:57] Speaker 04:
Thank you, Mr. Neelak.

[00:19:58] Speaker 04:
Mr. Rosenthal?

[00:20:03] Speaker 03:
Good morning, your honors.

[00:20:04] Speaker 03:
May it please the court, my name is Brian Rosenthal on behalf of the appellees.

[00:20:10] Speaker 03:
I want to start before I sort of get into some of those technical questions, which are great, just by emphasizing one thing.

[00:20:17] Speaker 03:
And that is that the focus of the one-on-one analysis has to be on the claims and not on the specification.

[00:20:25] Speaker 04:
But why don't you proceed for the moment on the theory that

[00:20:31] Speaker 04:
We reject your argument that the claims are not directed to error checking.

[00:20:36] Speaker 04:
And your argument is this just talks about the generation of data.

[00:20:40] Speaker 04:
But let's assume that it contemplates the generation of data for the purpose of error checking.

[00:20:46] Speaker 03:
Why are these three claims 101 and eligible?

[00:20:52] Speaker 03:
So we do not rely on the argument that they're divorced from error checking.

[00:20:59] Speaker 03:
They certainly, as they're construed for purposes of this motion, talk about generating additional data for use in error checking.

[00:21:06] Speaker 03:
So we have no problem with that.

[00:21:08] Speaker 03:
Why are they ineligible?

[00:21:10] Speaker 03:
The answer to that question is found by looking at the language of the claims themselves.

[00:21:14] Speaker 03:
And if you look at claim one, from which claims two through four depend, and really this is the beginning and end of the analysis as far as I'm concerned, there are three elements of the claim.

[00:21:29] Speaker 03:
There's a generating device that is configured to generate check data.

[00:21:33] Speaker 03:
And that has been, for purposes of this analysis, construed to mean generate supplemental data for use in error checking.

[00:21:42] Speaker 03:
Check data means, right.

[00:21:44] Speaker 03:
And check data, you know, we know what that means for use in error checking.

[00:21:48] Speaker 03:
Let's assume that that's correct.

[00:21:49] Speaker 03:
That's the argument that KPN is making.

[00:21:52] Speaker 03:
And then, this is very important, the varying device

[00:21:56] Speaker 03:
varies the original data and supplies that varied data to the generating device.

[00:22:01] Speaker 03:
Now claims two through four talk about how do you do that varying, but they don't otherwise change the structure of the claim.

[00:22:07] Speaker 04:
Well, but claim two adds something important, which is that the varying device modifies the permutation over time.

[00:22:17] Speaker 04:
That is, it uses different checks at different times for different blocks of data.

[00:22:22] Speaker 03:
But not different checks, but different permutations.

[00:22:24] Speaker 03:
Absolutely.

[00:22:25] Speaker 03:
It performs different permutations on the data at different times.

[00:22:29] Speaker 04:
But here's why.

[00:22:30] Speaker 04:
Which has the advantage of checking for systemic errors, right?

[00:22:35] Speaker 03:
Actually, there's nothing in the patent that talks about why that's an advantage at all.

[00:22:39] Speaker 03:
The only description in the patent about varying the permutation over time is contained in a single sentence at the bottom of column five and the top of column six.

[00:22:47] Speaker 03:
It says nothing about why that's useful, and I don't know why it's useful.

[00:22:51] Speaker 03:
Because when you're permutating the data,

[00:22:53] Speaker 03:
There's nothing in the patent that says that somehow the check data will be any different by permutating the data before you generate it.

[00:23:00] Speaker 03:
But the fundamental reason why these claims are ineligible is because there's no connection in the claim between the permutated data and the generated additional data.

[00:23:14] Speaker 03:
There's nothing in this claim that says or implies or has been construed to say that the very data is in any way used to generate the additional data.

[00:23:23] Speaker 03:
Did Stark say that?

[00:23:25] Speaker 03:
Yes, he did.

[00:23:25] Speaker 03:
He said that there's nothing in the claims that tells you how to use the reordered data in any way.

[00:23:32] Speaker 03:
And I can give you a citation for that, but it's in the middle of his opinion.

[00:23:36] Speaker 01:
What about where it says, I'm sorry.

[00:23:39] Speaker 02:
Oh, OK.

[00:23:41] Speaker 02:
I guess we have to, I mean, what you're asking for is a claim construction right now.

[00:23:47] Speaker 02:
I think you are, because there's a counterclaim construction that when you read this claim,

[00:23:52] Speaker 02:
in the context of this patent, the generating device configured to generate check data, the output that it's generating for check data is based on a certain input.

[00:24:04] Speaker 02:
What is that input?

[00:24:06] Speaker 02:
It's the input coming out of the varying device that varies the original data based on a permutation.

[00:24:13] Speaker 02:
And that permutation can change over time over the course of a series of different data blocks.

[00:24:21] Speaker 02:
When I understand your argument, which is I don't see actual words that say, take all of that varying data that you're generating in this claim and plug it into the generating device to generate said checked data.

[00:24:39] Speaker 02:
But the claim, at least the preamble, says you're producing error checking based on original data.

[00:24:47] Speaker 02:
And then this claim talks about taking original data,

[00:24:50] Speaker 02:
original data based on a permutation, and then ultimately a generating device to generate the checked data, I think there's a logic inside the claim that you're taking that original data, which is what is needed to generate checked data, and then you're doing all these manipulations based on different varying formulas to manipulate to generate the checked data.

[00:25:17] Speaker 02:
You're on it.

[00:25:20] Speaker ?:
I'm sorry.

[00:25:20] Speaker 02:
Let's just assume that that's a plausible reading of the claim.

[00:25:26] Speaker 02:
I understand your position, or at least one of your positions, is it's not a correct reading of the claim.

[00:25:32] Speaker 02:
But let's assume for the moment it is.

[00:25:35] Speaker 02:
Then where do we go?

[00:25:36] Speaker 03:
So first of all, the claim is still ineligible.

[00:25:41] Speaker 03:
And I'm not going to explain why.

[00:25:42] Speaker 03:
I do want to direct your attention to Appendix 656, which is where, at the oral argument,

[00:25:49] Speaker 03:
Council stated in response to a direct question from Judge Stark Do you need these claims to be construed and the answer was no?

[00:25:56] Speaker 03:
We don't need to get there and there is nothing in the language of the claims or in the construction That says the logical interpretation that you say is plausible.

[00:26:06] Speaker 03:
It may be a plausible interpretation we're not arguing claim construction here because no claim construction issue was ever raised and Council said that we're not arguing claim construction nobody argued that these claims could be interpreted as one twelve six paragraph No

[00:26:19] Speaker 02:
But my understanding is the entire premise of their argument, the other side's argument, has always been about the importance of varying the original data according to various formulas and then producing checked data based on that varied original data.

[00:26:40] Speaker 02:
So that is

[00:26:43] Speaker 02:
an articulation of a conception of what is going on inside these claims.

[00:26:48] Speaker 03:
You're right, Your Honor.

[00:26:50] Speaker 02:
And so I don't know, without doing a formal claim construction, why or how that particular conception of this claim, which to me doesn't sound crazy, should be dismissed out of hand.

[00:27:07] Speaker 03:
I don't disagree that it doesn't sound crazy.

[00:27:09] Speaker 03:
What I disagree is that KPN has ever argued that that's how the claim should be construed.

[00:27:12] Speaker 01:
But how are we supposed?

[00:27:14] Speaker 01:
Are we supposed to just not interpret the claim?

[00:27:16] Speaker 01:
Are we just not supposed to try to understand what it means as we make an assessment of whether the claim is eligible?

[00:27:23] Speaker 01:
I think that you're wrong.

[00:27:24] Speaker 01:
They probably didn't say, don't interpret this claim at all.

[00:27:27] Speaker 01:
I think they probably said you can give its plain and ordinary meaning, meaning that you don't have to interpret the claim.

[00:27:34] Speaker 01:
Right?

[00:27:34] Speaker 03:
Not really, but almost.

[00:27:36] Speaker 03:
What the parties actually said is that there were constructions which KPN sought and obtained in a different case against Samsung.

[00:27:43] Speaker 03:
And we said, let's use those constructions because they must be the most favorable.

[00:27:49] Speaker 03:
And in your opinion in content extraction,

[00:27:51] Speaker 03:
this Court has said it's appropriate to give the claims the most favorable meaning.

[00:27:57] Speaker 04:
How would those constructions bear on what we've just been talking about?

[00:28:01] Speaker 03:
Well, those constructions, when you read the language of those constructions, nowhere add to the claims the logical interpretation that Judge Chen was saying could be an interpretation that there's a connection between the reordered data and the generated additional data.

[00:28:15] Speaker 01:
Can I make sure I understand your position?

[00:28:17] Speaker 01:
I understand you to be saying that because

[00:28:20] Speaker 01:
The claim says a varying device to vary the original data prior to supplying it to the generating device.

[00:28:28] Speaker 01:
And it says there's a generating device configured to generate check data.

[00:28:32] Speaker 01:
Yes.

[00:28:33] Speaker 01:
That one of ordinary Scalmiard would not understand this claim to mean that that

[00:28:38] Speaker 01:
varied original data is the thing that the generating device is operating on to generate the check data.

[00:28:44] Speaker 03:
I'm not making that claim construction argument.

[00:28:46] Speaker 03:
What I'm saying is that nothing in the claims that's construed by the Samsung court, which all the parties have been operating under for this motion, says that.

[00:28:54] Speaker 01:
But new different litigations, new patent law issues that the courts are presented with

[00:29:03] Speaker 01:
might raise a new claim construction requirement, right?

[00:29:06] Speaker 01:
It might require that you have to understand a claim term.

[00:29:09] Speaker 04:
Sure.

[00:29:09] Speaker 04:
So, suppose we were to say, I know you don't like the notion that the

[00:29:16] Speaker 04:
claim would be construed the way Judge Chen was suggesting.

[00:29:19] Speaker 04:
But let's just hypothetically assume for the moment that we engage in that claim construction on appeal based on the intrinsic evidence and we reach that claim construction.

[00:29:30] Speaker 04:
Why under that claim construction are the claims two, three, and four patent ineligible?

[00:29:36] Speaker 03:
Very good.

[00:29:37] Speaker 03:
So under that claim construction, they're still ineligible because they do nothing more than take data, reorganize it,

[00:29:45] Speaker 03:
and then generate additional data.

[00:29:47] Speaker 03:
Now, the new data may well be based on the reordered data under this claim construction, right?

[00:29:54] Speaker 03:
But then they simply say it's for use in error checking.

[00:29:57] Speaker 03:
And the reason that that claim is still ineligible

[00:30:00] Speaker 03:
is for the same reason that the claim was ineligible in the content extraction case, or in Digitech, or in any of the other cases in which all the claim does is take data, reorganize it, put it in a different form, and then say use it.

[00:30:14] Speaker 04:
Let's take it one step further.

[00:30:16] Speaker 04:
Let's assume we reject that too, OK?

[00:30:19] Speaker 04:
And that what we say is that the claims here cover generating this permutation and using it for error checking, and that is required by claim two.

[00:30:30] Speaker 04:
that there are different iterations of the check data from one time period to another, okay?

[00:30:37] Speaker 04:
And that the benefit of that is that it enables to find systemic errors that you otherwise would.

[00:30:44] Speaker 04:
Let's assume that that's the case.

[00:30:47] Speaker 03:
Why is that patent ineligible?

[00:30:49] Speaker 03:
Well, because it's still, even with all of those assumptions, the claims actually never put to use the data that they generate.

[00:30:57] Speaker 03:
Now, I'm not saying that a patentee has to claim an entire system in order to have a concrete eligible invention.

[00:31:04] Speaker 04:
However, the thing that the patentee... Suppose the claims, let's say hypothetically, the claims did tell you to put the data to use for error checking by varying it from one time period to another.

[00:31:17] Speaker 04:
in order to check for systemic error.

[00:31:20] Speaker 04:
Let's suppose that this thing were clearly written to say all of those things.

[00:31:24] Speaker 04:
Would that be patent ineligible?

[00:31:26] Speaker 03:
I think if the claims were written to cover what's actually described as the invention in the patent, which is doing what you just said on both sides of a transmission and then comparing the results to determine whether an error has occurred,

[00:31:39] Speaker 03:
I would not be standing up here with the same argument.

[00:31:41] Speaker 02:
I guess what here's my concern with that response is that it seems intuitive that you ought to be able to patent an improvement to a particular device in an overall system rather than be forced to go further and claim the entire system with one of the claim elements being the improved device.

[00:32:07] Speaker 02:
And so here,

[00:32:09] Speaker 02:
Why isn't there an argument that this check data generator device, that's what the claim is directed to, is an improved check data generator device in the sense that this particular check data generator device can now give you

[00:32:29] Speaker 02:
or give the system an important new tool to uncover systemic error through the variance of the permutation formula.

[00:32:38] Speaker 02:
And so thanks to the power of software, what we have is a device that's something of a shape-shifter.

[00:32:44] Speaker 02:
At block one, it uses permutation formula one.

[00:32:49] Speaker 02:
At block seven, it's going to use permutation formula number seven, and so on and so on in such a way that now you're going to

[00:32:58] Speaker 02:
figure out the systemic error that keeps repeating through all of the data blocks.

[00:33:02] Speaker 02:
So what we've got here is arguably a shape-shifting check data generator.

[00:33:08] Speaker 03:
Your Honor, if they had claimed that dynamic error check generator and

[00:33:15] Speaker 03:
in some kind of an application, we would be dealing with a different case.

[00:33:18] Speaker 02:
What does that mean by in some kind of application?

[00:33:20] Speaker 02:
Because why can't you just claim the features of your improved device, full stop period, don't go further and say, and then now I'm going to convert my device claim into a method claim because now I'm going to then recite all the steps of how

[00:33:37] Speaker 02:
this device gets used in an overall system.

[00:33:39] Speaker 03:
You absolutely can do it as long as the thing that you are claiming is itself a concrete invention.

[00:33:45] Speaker 03:
And if I may analogize, I think it was Diamond v. Deer case, long time ago, was the case about using a mathematical formula for curing rubber.

[00:33:54] Speaker 03:
And in that case, the Supreme Court said,

[00:33:57] Speaker 03:
that that was a patent eligible claim because they talked about how you use the mathematical formula in the process of curing rubber.

[00:34:07] Speaker 03:
Now, if they had instead written a claim that said, cure rubber using this mathematical formula,

[00:34:14] Speaker 03:
I think we would all agree that that is not a patent-eligible invention, because they have not described how to use this abstract idea, other than simply say, put it in this context.

[00:34:25] Speaker 01:
What if, though, we think that the claim does describe how you generate the error check data?

[00:34:33] Speaker 01:
It doesn't explain how you do the comparison at the end, but it does explain how you generate the error check data.

[00:34:40] Speaker 03:
To me, that is not concrete enough under this court's precedent.

[00:34:43] Speaker 03:
Because under this court's precedent, Digitech content extraction, the Capital One financial case, the Intellectual Ventures case, and several others, all of those cases talk about taking data, manipulating the data, generating additional data, putting it in a new form.

[00:35:06] Speaker 03:
I'm reminded of the court's decision in Digitech.

[00:35:09] Speaker 03:
And in that case, this is what they said.

[00:35:12] Speaker 03:
Nothing in the claim language expressly ties the method to an image processor.

[00:35:16] Speaker 03:
This was about creating a device profile.

[00:35:19] Speaker 03:
Let me talk about the facts for a moment.

[00:35:22] Speaker 03:
In that case, there was a claim that said, take all of this data from one source, take all of this data from another source, combine them into a device profile.

[00:35:31] Speaker 03:
And the specification talked about how important that was in doing image processing.

[00:35:36] Speaker 03:
This court said that the claim generically recites a process of combining the two data sets into a device profile.

[00:35:43] Speaker 03:
It does not claim the processor's use of the profile in the capturing, transforming, or rendering of a digital image.

[00:35:51] Speaker 03:
And the point that the court was making is the same point that we are making here, which is that just explaining, if the court agrees.

[00:35:58] Speaker 04:
What sounds to me is that what you're saying is that there is a patent eligible invention here, but they did a bad job.

[00:36:05] Speaker 04:
of writing their patent claims to cover it.

[00:36:07] Speaker 03:
I couldn't have said it better, except I would probably say there may be a patent-eligible invention here, because I'm a lawyer.

[00:36:14] Speaker 03:
But I totally agree, Your Honor, that had they claimed something that is actually what's described as the innovation, as the technological solution, had they done that, we would be in a very different situation.

[00:36:29] Speaker 01:
And one of the things that might be different between this case and the case that you just described is that error check data has a very well-known meaning to one of ordinary scale in the earth.

[00:36:40] Speaker 03:
Well, error check data may well have a well-known meaning, but the only thing that this is doing, so too does a device profile in the Digitech case.

[00:36:49] Speaker 01:
I don't know if device profile has a well-known specific meaning.

[00:36:53] Speaker 01:
Device profile could have very different meanings depending on what device you're talking about.

[00:36:57] Speaker 01:
My refrigerator is a device that could have a profile.

[00:37:00] Speaker 01:
On the other hand, error check data is something that always relates to a communication field.

[00:37:07] Speaker 01:
And the sending of data from one point

[00:37:10] Speaker 01:
to a destination and seeing whether there's error in that communication.

[00:37:14] Speaker 01:
It's pretty narrow.

[00:37:16] Speaker 03:
Well, let me take as a given that that's true.

[00:37:18] Speaker 03:
I'm not sure I agree with that, but even if that were true, to me, the answer doesn't change as to eligibility, because all you have done with this claim is describe, and it's a device, but that doesn't do anything.

[00:37:30] Speaker 03:
It's a method that is implemented on generic hardware to do the following.

[00:37:35] Speaker 03:
take input data, reorder it, generate additional data for use in error checking.

[00:37:41] Speaker 03:
How do you use that error check data?

[00:37:43] Speaker 03:
Should you be doing the comparisons against both?

[00:37:46] Speaker 03:
All of that stuff is absolutely absent from the claim.

[00:37:49] Speaker 03:
It is simply generating data, and the language that they hang their hat on, that it's for use in error checking, that does no more than place this in a technological context, which the Supreme Court said in Alice was insufficient.

[00:38:04] Speaker 03:
It's not sufficient to simply say, when you generate this data, use it for error checking somehow.

[00:38:10] Speaker 03:
That's not sufficient, according to Alice, to simply say do it in some environment.

[00:38:15] Speaker 03:
I'm not sure what my time situation is.

[00:38:17] Speaker 03:
I think you're out of time.

[00:38:18] Speaker 03:
I am.

[00:38:20] Speaker 03:
All right.

[00:38:21] Speaker 03:
Thank you very much.

[00:38:21] Speaker 04:
Mr. Healy, we'll give you a couple minutes, two minutes.

[00:38:30] Speaker 00:
I want to push back on the last point that Mr. Rosenthal just said, which is that the description for use in error checking is just an intended use.

[00:38:36] Speaker 00:
We specifically incorporated that claim construction in the body of the claim to describe the characteristics of the data that needed to be produced.

[00:38:43] Speaker 04:
It's not very well written.

[00:38:44] Speaker 00:
These claims are not really well written, right?

[00:38:47] Speaker 00:
Your Honor, I was not involved in the prosecution.

[00:38:49] Speaker 00:
I certainly could agree that they could be written in a better way, perhaps.

[00:38:54] Speaker 00:
They're originally in Dutch, and I think there may be a translation issue there.

[00:38:57] Speaker 00:
But the core point is this is not an intended use of this statement.

[00:39:01] Speaker 00:
This describes the specific output that's required

[00:39:05] Speaker 00:
which makes it a more concrete invention.

[00:39:07] Speaker 00:
I also want to talk, just respond, the advantage of the varying device.

[00:39:14] Speaker 00:
Your Honor, I think your analogy is apt.

[00:39:16] Speaker 00:
If KPN had invented a better brake pad, one that had better, withstood temperatures better, had longer longevity, the fact that it didn't claim the car to incorporate that brake pad in, the fact that it didn't incorporate stopping the car, does not mean that the invention that was claimed as concrete

[00:39:35] Speaker 00:
Specific device that's contemplated certainly to be incorporated into a larger system or process or vehicle in order to realize that benefit It doesn't mean that it's not a better Permutating device and that's what exactly is claimed here And I think with respect to the the benefit to be realized by this again the specification makes clear that The benefit is specific to the permutating device.

[00:40:00] Speaker 00:
It's not there's not this overarching need to incorporate it into a larger

[00:40:04] Speaker 00:
process in order to absolutely realize that and again I would refer the court to appendix 71 columns 2 lines 48 through 58 says expressly the invention is also based on the insight that a variable checking function can almost always prevent the non-detection of repetitive errors a further aspect of the invention is based on the insight that the varying of the checking function is

[00:40:27] Speaker 00:
can be advantageously based on the data to be checked.

[00:40:31] Speaker 00:
That's claim three.

[00:40:32] Speaker 00:
And that the varying of the data can be used to accomplish a varying, i.e., time-dependent checking function.

[00:40:36] Speaker 00:
That's claim two.

[00:40:38] Speaker 00:
Mr. Healy, I think we're out of time.

[00:40:40] Speaker 04:
Thank you.

[00:40:41] Speaker 04:
Thank you, Your Honor.

[00:40:41] Speaker 04:
I thank both counsel, the cases submitted.